SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2406.16976 · ICLR · 2025

Efficient Evolutionary Search Over Chemical Space with Large Language Models

Lingkai Kong, Chao Zhang, Yue Yu, Wenhao Gao, Yanqiao Zhu, Yuanqi Du, Haorui Wang, Kirill Neklyudov, Yuchen Zhuang, Felix Strieth-Kalthoff, Marta Skreta, Alán Aspuru-Guzik, and 2 more

arXiv · PDF · Open in the Atlas

Code that ran

5 of 6 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 2 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 3 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2406.16976.json · badge: /api/ran/2406.16976.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
zoom-wang112358/MOLLEO canonical (classifier label)
classifier basis: census
5 ran · 1 no recorded run
FunctionStatusWhere it lives
get_device Ran
ran; our contract draft was wrong, not the code
zoom-wang112358/MOLLEO/multi_objective/main/molleo_multi/network.py
code served (permissive licence) · get_code("3e00ba37467a39b3")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned str. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

ARGS = {
    'use_gpu': False
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_lr Ran
ran, honoured the contract we drafted
zoom-wang112358/MOLLEO/multi_objective/main/molleo_multi/mol_lm.py
code served (permissive licence) · get_code("ea6c82ed6a138df9")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Sonnet 4.5, prompt python_fixture_v1) and returned float, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import math

# The function get_lr takes:
# - t: a float representing the current time/progress (typically between 0 and 1)
# - initial_lr: a float representing the initial learning rate
# - rampdown: a float representing the rampdown fraction (default 0.25)
# - rampup: a float representing the rampup fraction (default 0.05)

ARGS = {
    't': 0.3,
    'initial_lr': 0.001,
    'rampdown': 0.25,
    'rampup': 0.05
}
Setup block drafted by
a language model, Claude Sonnet 4.5, prompt python_fixture_v1
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
mean_pooling Ran
ran; our contract draft was wrong, not the code
zoom-wang112358/MOLLEO/multi_objective/main/molleo_multi/mol_lm.py
code served (permissive licence) · get_code("3ca541ab8bce64d7")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 2, dtype family float, shape (3, 4) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# token_embeddings: shape (batch_size, sequence_length, embedding_dim)
# Let's use batch=2, sequence_length=3, embedding_dim=4
token_embeddings = torch.randn(2, 3, 4)

# attention_mask: boolean tensor of shape (batch_size, sequence_length)
# The function uses ~attention_mask (bitwise NOT), so it expects boolean
attention_mask = torch.tensor([
    [True, True, False],
    [True, False, False]
])

ARGS = {
    "token_embeddings": token_embeddings,
    "attention_mask": attention_mask
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (3, 4) — shape (3, 4) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
padarray Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
zoom-wang112358/MOLLEO/multi_objective/main/molleo_multi/mol_lm_utils.py
code served (permissive licence) · get_code("edaa7e0058da6c0a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned ndarray, ndim 1, dtype family int, shape (7,) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

# Analyze the function:
# - A: array-like input (used with len())
# - size: integer specifying target size
# - value: scalar for padding (default 0)

# Create a small array with 3 elements
A = np.array([1, 2, 3])

# Set target size to 7 (larger than len(A) which is 3)
size = 7

# Set padding value to -1
value = -1

ARGS = {
    "A": A,
    "size": size,
    "value": value
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type ndarray, ndim 1, dtype family int, shape (7,) — shape (7,) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
preprocess_each_sentence Ran
ran, honoured the contract we drafted
zoom-wang112358/MOLLEO/multi_objective/main/molleo_multi/mol_lm_utils.py
code served (permissive licence) · get_code("dbb8233cd18cc6c4")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned list, ndim 2, dtype family int; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import numpy as np

class MockTokenizer:
    def __call__(self, sentence, truncation=False, max_length=None, padding=None, return_tensors=None):
        # Return mock tokenizer output with input_ids and attention_mask
        # Shape should match what the function expects
        batch_size = 1
        seq_len = max_length if max_length else 10
        input_ids = np.ones((batch_size, seq_len), dtype=np.int64)
        attention_mask = np.ones((batch_size, seq_len), dtype=np.int64)
        return {
            'input_ids': input_ids,
            'attention_mask': attention_mask
        }

tokenizer = MockTokenizer()
sentence = "This is a test sentence."
max_seq_len = 10

ARGS = {
    'sentence': sentence,
    'tokenizer': tokenizer,
    'max_seq_len': max_seq_len
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 2, dtype family int — a list output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_bond_indeces Not yet run
unverified (harvested, no recorded run)
zoom-wang112358/MOLLEO/multi_objective/main/molleo_multi/features.py
code served (permissive licence) · get_code("1257979c9fb38df8")
How it ran

no invocation recorded.

Also present in other papers' code

Of 6 samples lifted from this paper's linked repositories, 6 carry a function identity and 0 do not. Of the 6 distinct functions with an identity, 3 are also present in other papers whose code we hold (3 with at least one paper dated earlier, 0 the same month, 0 later) and 3 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
get_lr 7 other papers
7 dated earlier
1912.04958 (2019-12)
Analyzing and Improving the Image Quality of StyleGAN
2101.06006 (2021-01)
The Geometry of Deep Generative Image Models and its Applications
2102.06108 (2021-02)
SWAGAN: A Style-based Wavelet-driven Generative Model
and 4 more not named here
get_bond_indeces 1 other paper
1 dated earlier
2106.04011 (2021-06)
JANUS: Parallel Tempered Genetic Algorithm Guided by Deep Neural Networks for Inverse Molecular Design
get_device 1 other paper
1 dated earlier
2106.04011 (2021-06)
JANUS: Parallel Tempered Genetic Algorithm Guided by Deep Neural Networks for Inverse Molecular Design

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Molecular discovery, when formulated as an optimization problem, presents significant computational challenges because optimization objectives can be nondifferentiable. Evolutionary Algorithms (EAs), often used to optimize black-box objectives in molecular discovery, traverse chemical space by performing random mutations and crossovers, leading to a large number of expensive objective evaluations. In this work, we ameliorate this shortcoming by incorporating chemistryaware Large Language Models (LLMs) into EAs. Namely, we redesign crossover and mutation operations in EAs using LLMs trained on large corpora of chemical information. We perform extensive empirical studies on both commercial and open-source models on multiple tasks involving property optimization, molecular rediscovery, and structure-based drug design, demonstrating that the joint usage of LLMs with EAs yields superior performance over all baseline models across single-and multi-objective settings. We demonstrate that our algorithm improves both the quality of the final solution and convergence speed, thereby reducing the number of required objective evaluations.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2406.16976")
get_code_for_paper("2406.16976")
have("2406.16976")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2406.16976.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2406.16976.svg)](https://syntology.ai/paper/2406.16976)

Connect an agent — have() is free.